Financial escort vertical type large model robot fusing multiple modes

Abstract

The invention belongs to the technical field of robots, and discloses a multi-modal fused financial escort vertical type large model robot, which comprises a multi-modal sensing layer module, a data warehouse module, a financial escort vertical type large model module and a business support application module, according to the invention, through fusion of the multi-modal data and accurate identification of the financial escort vertical large model, various safety risks such as false claim of personnel and theft of bags and suitcases in the escort process can be effectively prevented, and safety and reliability of financial escort are ensured. According to the system, automation and informatization of escort registration and management are achieved, complexity and errors of manual operation are reduced, and working efficiency is improved. Meanwhile, data analysis and prediction functions provided by the financial escort vertical class large model can provide a scientific basis for scheduling and management of escort tasks, and the management process is further optimized. And the real-time monitoring and abnormity alarm functions enable monitoring personnel to find problems in time and take measures, so that the timeliness and effectiveness of emergency processing are improved.

Description

Technical Field

[0001] The present invention belongs to the field of robot technology, and in particular relates to a multi-modal financial escort vertical large-scale model robot. Background Art

[0002] In banks, securities and other financial institutions, the escort and handover registration system is widely used to ensure the safe escort and handover registration of cash, bills, valuables, etc.

[0003] The most advanced escort handover systems currently use the following technical solutions:

[0004] Personnel information management: Manage the name, gender, age, company department and position of escort personnel through structured data. Through fingerprint recognition, face recognition, etc., make the identity verification of handover personnel more accurate and reliable, and effectively prevent risks such as impersonation.

[0005] Vehicle trajectory management: With the help of video surveillance technology and Internet of Things technology, the entire process of escort handover is captured and monitored in real time, and the video data is uploaded to the server so that back-end managers can view and trace back at any time, thereby enhancing the supervision of the escort process.

[0006] With the rapid development of artificial intelligence (AI) technology, the status of general-purpose large models in AI development and application is becoming increasingly prominent. General-purpose large models refer to large-scale and powerful deep learning models that can handle a variety of different tasks and data types. It usually learns various features and patterns by training a large amount of data, thereby achieving fast and accurate processing of new data.

[0007] As a training ground, the general large model provides a strong foundation for various AI applications, supporting the development of applications in many fields, from speech recognition to image recognition, from natural language processing to recommendation systems. Currently, the general large model mainly plays a role in the following fields:

[0008] 1. Speech recognition: The application of the general large model in the field of speech recognition has achieved remarkable results. By training a large amount of speech data, the general large model can achieve accurate recognition and understanding of various languages.

[0009] 2. Image recognition: In the field of image recognition, general large models also play an important role. By training a large amount of image data, general large models can recognize information such as objects, scenes, and emotions in images.

[0010] 3. Natural language processing: Natural language processing is another important application area of ​​the general large model. By training a large amount of text data, the general large model can achieve a variety of tasks such as text classification, sentiment analysis, and machine translation.

[0011] 4. Recommendation system: In the field of recommendation system, the general large model can recommend relevant content and services to users based on their historical behaviors and preferences. This can not only improve user experience, but also create more commercial value for enterprises.

[0012] Through the above analysis, the problems and defects of the prior art are as follows:

[0013] When the general big model is combined with the financial escort industry, the material dimensions are still not rich enough. It is difficult for the general big model to understand the characteristics and rules of the financial escort business only by mining and analyzing the existing structured business data. Summary of the invention

[0014] In response to the problems existing in the prior art, the present invention provides a large-scale vertical model robot for financial escort that integrates multiple modes.

[0015] The present invention is implemented as follows: a multi-modal financial escort vertical large model robot includes:

[0016] Multimodal perception layer module, data warehouse module, financial escort vertical large model module, business support application module;

[0017] The multimodal perception layer module is connected with the data warehouse module and the supporting application module to obtain biometric data, location data, luggage data, and visual data;

[0018] The data warehouse module is connected with the multimodal perception layer module, the financial escort vertical model module, and the business support application module to store data;

[0019] The large vertical model module of financial escort is connected with the data warehouse module and the business support application module for data preprocessing and feature extraction, personnel and luggage identification, abnormal behavior detection, and data prediction;

[0020] The business support application module is connected to the multimodal perception layer module, the data warehouse module, and the financial escort vertical model module, and is used for personnel registration, luggage registration, task scheduling, monitoring and alarm, and data management and query.

[0021] Furthermore, the multimodal perception layer module:

[0022] (1) Biometric data: Use fingerprint recognition and iris recognition biometric technology to collect fingerprint and iris information of escort personnel;

[0023] (2) Location data: GPS positioning technology is used to obtain the location information of the escort vehicle in real time. Combined with the geographic information system, the driving route and real-time location of the escort vehicle can be accurately tracked so that the monitoring center can keep track of the escort progress at any time and detect abnormal situations in a timely manner.

[0024] (3) Luggage data: Sensors and NFC identification chips are installed in the escorted cash boxes. The robot collects the unique ID of the luggage, as well as the weight, temperature, humidity status data of the luggage, and whether it has been illegally opened, to ensure the safety and integrity of the luggage during the escort process;

[0025] (4) Visual data: When arriving at the escort handover site, at key locations such as the safe door and bank branch entrance, the robot’s high-definition camera collects the facial information of the vault guards and the site environment information in real time to register the escort receipt behavior.

[0026] Furthermore, the financial escort vertical model module:

[0027] (1) Data preprocessing and feature extraction: Preprocess the collected multimodal data to identify and extract valuable features;

[0028] (2) Personnel and luggage identification: Train the classification and recognition capabilities of the financial escort vertical large model to accurately match the collected biometric features of personnel and visual features of luggage. Through a large amount of training data, accurately determine whether the personnel are legal business outlets’ warehouse managers, vault managers, cash delivery personnel, and escorts, and whether the escorted luggage is the registered target luggage, effectively preventing the occurrence of false or wrong collection.

[0029] (3) Abnormal behavior detection: Based on the learning of a large amount of normal escort behavior data, the escort personnel’s behavior and actions are analyzed and monitored in real time to identify abnormal behavior patterns;

[0030] (4) Data prediction: By analyzing the historical data, real-time road conditions, and weather conditions of the escort routes, we can discover the underlying patterns and trends, predict possible risk points and congested sections, and plan and schedule escort tasks in advance.

[0031] Furthermore, the business support application module:

[0032] (1) Personnel registration module: registers and manages the basic information and biometric information of escort personnel at various positions; when personnel join the service, their facial, fingerprint, and iris biometrics are collected and bound to their personal identity information and stored in the system database; during escort missions, identity verification of personnel at various positions is performed through visual and biometric recognition technology to ensure that only legitimate personnel participate in escort missions;

[0033] (2) Luggage registration module: Detailed information about the escorted luggage is registered, including the luggage number, type, weight, and bank branch information, and associated with the corresponding escort task; during the escort process, the luggage is monitored in real time through visual recognition and sensor data to ensure that the status of the luggage is consistent with the registration information;

[0034] (3) Task Scheduling Module: Develop a reasonable escort task plan based on the bank’s business needs and the availability of escort resources; optimize and adjust escort routes and time arrangements based on the data analysis and prediction results provided by the financial escort vertical model;

[0035] (4) Monitoring and alarm module;

[0036] (5)Data management and query module.

[0037] Furthermore, the monitoring and alarm module: receives and displays information from multimodal data acquisition equipment in real time, and conducts all-round monitoring of the escort process; when the financial escort vertical large model detects an abnormal situation, the system automatically sends an alarm signal and pushes relevant information to monitoring personnel and managers.

[0038] Furthermore, the data management and query module is responsible for storing, managing and querying various types of data generated in the system, including personnel information, luggage information, escort task records, and surveillance video data, and supports query and statistical analysis according to different conditions, providing data support and decision-making basis for escort management.

[0039] Another object of the present invention is to provide a multi-modal financial escort vertical large model robot control method, including:

[0040] Step 1, obtaining biometric data, location data, luggage data, and visual data through a multimodal perception layer module;

[0041] Step 2, storing data through the data warehouse module;

[0042] Step 3: Data preprocessing and feature extraction, personnel and luggage recognition, abnormal behavior detection, and data prediction through the large model module of financial escort vertical category;

[0043] Step 4: Use the business support application modules for personnel registration, luggage registration, task scheduling, monitoring and alarm, and data management and query.

[0044] Another object of the present invention is to provide a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the multi-modal financial escort vertical large-model robot control method.

[0045] Another object of the present invention is to provide a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor executes the steps of the multi-modal integrated financial escort vertical large-scale model robot control method.

[0046] Another object of the present invention is to provide an information data processing terminal, which is used to realize the multi-modal integrated financial escort vertical large-scale model robot.

[0047] In combination with the above technical solutions and the technical problems solved, please analyze the advantages and positive effects of the technical solutions to be protected by the present invention from the following aspects:

[0048] First, in view of the technical problems existing in the above-mentioned prior art and the difficulty of solving the problems, the technical solutions to be protected by the present invention and the results and data during the research and development process are closely combined to analyze in detail and deeply how the technical solutions of the present invention solve the technical problems, and some creative technical effects brought about after solving the problems. The specific description is as follows:

[0049] The financial escort field has extremely high requirements for security, accuracy and efficiency, and it is difficult to further improve them by relying solely on existing traditional technologies. At the same time, AI technology has made rapid progress, but there is a lack of general large-scale model industry application scenarios in the field of financial escort.

[0050] In addition to traditional structured data, the financial escort vertical model also needs to integrate more unstructured data, such as video surveillance data, voice communication data, geographic information data, etc. Through the fusion and analysis of multimodal data, we can have a more comprehensive understanding of the situation at the escort site and provide stronger support for decision-making.

[0051] The vertical large model processes and analyzes massive amounts of data quickly, thereby more accurately identifying potential risk factors, such as robbery risk, traffic accident risk, personnel violation risk, etc. At the same time, combined with machine learning algorithms, the probability and impact of risks are evaluated, and corresponding risk response strategies are formulated to help financial escort companies achieve more refined risk management and control, and improve risk assessment and management capabilities.

[0052] The present invention realizes a multi-modal financial escort vertical large model robot, which integrates data from multiple modes and utilizes the powerful analysis ability of the vertical large model to achieve efficient and accurate registration and management of various types of information in the financial escort process, thereby improving the security and management efficiency of financial escort.

[0053] The financial escort robot is placed on the escort vehicle in its normal working state and runs with the escort vehicle.

[0054] Improve security: Through the fusion of multimodal data and the accurate identification of the large vertical model of financial escort, various security risks in the escort process, such as people impersonating and luggage theft, can be effectively prevented, ensuring the safety and reliability of financial escort.

[0055] Improve management efficiency: The system realizes the automation and informatization of escort registration and management, reduces the tediousness and errors of manual operation, and improves work efficiency. At the same time, the data analysis and prediction functions provided by the financial escort vertical model can provide a scientific basis for the dispatch and management of escort tasks and further optimize the management process.

[0056] Enhanced emergency response capabilities: Real-time monitoring and abnormal alarm functions enable monitoring personnel to promptly identify problems and take measures, improving the timeliness and effectiveness of emergency response. The financial escort vertical model can also provide decision support for emergency response based on historical data and real-time conditions, reducing the impact of emergencies on escort work.

[0057] Data value mining: The large amount of multimodal data accumulated by the system has important value. Through the mining and analysis of the financial escort vertical model, it can provide financial institutions with more meaningful information, such as risk assessment, business optimization, etc., and promote the sustainable development and innovation of financial escort business.

[0058] The traditional financial escort handover registration system is transformed and upgraded, introducing multimodal data sources such as video, biometrics, GPS information, and NFC identification. The escort is made safer and more efficient by training a large vertical model of financial escort.

[0059] Second, as auxiliary evidence of the inventiveness of the claims of the present invention, it is also reflected in the fact that the technical solution of the present invention solves the technical problems that people have always been eager to solve but have never been successful:

[0060] Improved security: Traditional financial escort systems often rely on manual monitoring and physical security measures, which makes it difficult to completely prevent internal violations or external security threats. By introducing multimodal data fusion (such as video surveillance, biometrics, etc.) and specially trained large models, more accurate identification of abnormal behaviors during the escort process can be achieved, thereby greatly improving security.

[0061] Improved management efficiency: In the past, the registration and management of financial escorts usually involved a lot of manual operations, which not only increased the possibility of errors, but also reduced work efficiency. The use of automated and information-based management systems, combined with the data analysis capabilities of vertical large models, can significantly reduce manual intervention, optimize work processes, and complete tasks more efficiently and accurately.

[0062] Enhanced emergency response capabilities: Traditional systems react slowly to emergencies and lack scientific basis for decision-making. However, real-time monitoring, abnormal alarms, and intelligent decision support based on historical data and real-time conditions can make emergency response faster and more effective, ensuring that a high level of operation can be maintained even in emergency situations.

[0063] Data value mining: In the past, a large amount of data generated during financial escort was not fully utilized. Now, by integrating multiple types of data and applying advanced analytical techniques, valuable information can be extracted from these data to provide better services and support for financial institutions, while also promoting business development and innovation. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 This is a structural block diagram of a large-scale vertical model robot for financial escort that integrates multiple modes provided in an embodiment of the present invention.

[0065] Figure 2 It is a flow chart of the financial escort vertical large model module method provided by an embodiment of the present invention.

[0066] Figure 3 It is a flow chart of the control method of a large-scale vertical model robot for financial escort integrating multiple modes provided in an embodiment of the present invention.

[0067] Figure 4 It is a diagram of the overall system architecture provided by an embodiment of the present invention.

[0068] Figure 5 It is a key indicator analysis diagram before and after the use of the present invention provided by the embodiment of the present invention.

[0069] Figure 1 In: 1. Multimodal perception layer module; 2. Data warehouse module; 3. Financial escort vertical large model module; 4. Business support application module. DETAILED DESCRIPTION

[0070] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0071] like Figure 1 As shown, a multi-modal financial escort vertical large model robot provided by an embodiment of the present invention includes:

[0072] Multimodal perception layer module 1, data warehouse module 2, financial escort vertical large model module 3, business support application module 4.

[0073] The multimodal perception layer module 1 is connected to the data warehouse module 2 and the supporting application module 4 to obtain biometric data, location data, luggage data, and visual data;

[0074] The data warehouse module 2 is connected with the multimodal perception layer module 1, the financial escort vertical model module 3, and the business support application module 4 to store data;

[0075] The financial escort vertical model module 3 is connected to the data warehouse module 2 and the business support application module 4, and is used for data preprocessing and feature extraction, personnel and luggage identification, abnormal behavior detection, and data prediction;

[0076] The business support application module 4 is connected with the multimodal perception layer module 1, the data warehouse module 2, and the financial escort vertical model module 3, and is used for personnel registration, luggage registration, task scheduling, monitoring and alarm, data management and query.

[0077] The multimodal perception layer module provided by the embodiment of the present invention:

[0078] (1) Biometric data: Fingerprint and iris recognition biometrics are used to collect fingerprint and iris information of escort personnel as an important basis for personnel identity verification, ensuring that only authorized personnel can participate in escort work and preventing identity fraud.

[0079] (2) Location data: GPS positioning technology is used to obtain the location information of the escort vehicle in real time. Combined with the Geographic Information System (GIS), the escort vehicle’s driving route and real-time location can be accurately tracked so that the monitoring center can keep track of the escort progress at any time and promptly detect abnormal situations, such as deviation from the scheduled route.

[0080] (3) Luggage data: Sensors and NFC identification chips are installed in the escorted cash boxes. The robot collects the unique ID of the luggage, as well as the weight, temperature, humidity status data of the luggage, and whether it has been illegally opened, to ensure the safety and integrity of the luggage during the escort process;

[0081] (4) Visual data: When arriving at the escort handover site, such as at key locations such as the entrance to a safe or a bank branch, the robot’s high-definition camera collects real-time facial information of the safekeeping personnel and site environment information to register the escort receipt behavior.

[0082] like Figure 2 As shown, the financial escort vertical model module provided by the embodiment of the present invention is:

[0083] S101, data preprocessing and feature extraction:

[0084] Pre-process the collected multimodal data to identify and extract valuable features; for example, accurately identify the shape, color, and texture features of escorted bags from visual images, and extract unique identity feature vectors from biometric data, providing a basis for subsequent identification and analysis;

[0085] S102, Personnel and Baggage Identification:

[0086] Train the classification and recognition capabilities of the financial escort vertical model to accurately match the collected biometric features of personnel and visual features of luggage; through a large amount of training data, accurately determine whether the personnel are legal business outlets, vault managers, cash collectors and escorts, and whether the escorted luggage is the registered target luggage, effectively preventing the occurrence of false or wrong collection;

[0087] S103, abnormal behavior detection:

[0088] Based on the learning of a large amount of normal escort behavior data, the escort personnel's behavior and actions are analyzed and monitored in real time to identify abnormal behavior patterns, such as personnel wandering in non-specified areas and bags being moved abnormally. Once an abnormality is found, an alarm is immediately issued to remind the monitoring personnel to deal with it in time;

[0089] S104, data prediction:

[0090] By analyzing the historical data, real-time road conditions, weather conditions and other multi-dimensional information of the escort routes, we can discover the potential patterns and trends, predict possible risk points and congested sections, plan and schedule escort tasks in advance, arrange safer and faster routes for escort tasks, reduce risks and improve efficiency, and improve the efficiency and safety of escort work.

[0091] The business support application module provided by the embodiment of the present invention:

[0092] (1) Personnel registration module: registers and manages the basic information and biometric information of escort personnel at various positions; when personnel join the service, their facial, fingerprint, and iris biometrics are collected and bound to their personal identity information and stored in the system database; during escort missions, identity verification of personnel at various positions is performed through visual and biometric recognition technology to ensure that only legitimate personnel participate in escort missions;

[0093] (2) Luggage registration module: Detailed information about the escorted luggage is registered, including the luggage number, type, weight, and bank branch information, and associated with the corresponding escort task; during the escort process, the luggage is monitored in real time through visual recognition and sensor data to ensure that the status of the luggage is consistent with the registration information;

[0094] (3) Task Scheduling Module: Develop a reasonable escort task plan based on the bank's business needs and the availability of escort resources; optimize and adjust escort routes and schedules based on the data analysis and prediction results provided by the financial escort vertical model to improve escort efficiency and reduce risks; at the same time, track the execution of escort tasks in real time and handle abnormal situations in tasks in a timely manner;

[0095] (4) Monitoring and alarm module: It receives and displays information from multi-modal data acquisition equipment in real time and monitors the escort process in all directions. When the financial escort vertical model detects an abnormal situation, the system automatically sends an alarm signal and pushes relevant information to monitoring personnel and management personnel so that timely measures can be taken to ensure escort safety.

[0096] (5) Data management and query module: Responsible for storing, managing and querying various types of data generated in the system, including personnel information, luggage information, escort mission records, and surveillance video data. It supports query and statistical analysis based on different conditions, providing data support and decision-making basis for escort management.

[0097] like Figure 3 As shown, a multi-modal financial escort vertical large model robot control method provided by an embodiment of the present invention includes:

[0098] S201, acquiring biometric data, location data, luggage data, and visual data through a multimodal perception layer module;

[0099] S202, storing data through a data warehouse module;

[0100] S203, through the data preprocessing and feature extraction of the financial escort vertical large model module, personnel and luggage recognition, abnormal behavior detection, and data prediction;

[0101] S204, through the business support application modules personnel registration, luggage registration, task scheduling, monitoring and alarm, data management and query.

[0102] Another object of the present invention is to provide a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the multi-modal financial escort vertical large-model robot control method.

[0103] Another object of the present invention is to provide a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor executes the steps of the multi-modal integrated financial escort vertical large-scale model robot control method.

[0104] Another object of the present invention is to provide an information data processing terminal, which is used to realize the multi-modal integrated financial escort vertical large-scale model robot.

[0105] The present invention is specifically implemented:

[0106] The present invention realizes a multi-modal financial escort vertical large model robot, which integrates data from multiple modes and utilizes the powerful analysis ability of the vertical large model to achieve efficient and accurate registration and management of various types of information in the financial escort process, thereby improving the security and management efficiency of financial escort.

[0107] The financial escort robot is placed on the escort vehicle in its normal working state and runs with the escort vehicle.

[0108] The robot's appearance is not limited, and its built-in system architecture Figure 4 :

[0109] 1. Multimodal perception layer

[0110] (1) Biometric data: Use biometric technologies such as fingerprint recognition and iris recognition to collect fingerprints, irises and other information of escort personnel as an important basis for personnel identity verification, ensuring that only authorized personnel can participate in escort work and preventing identity fraud.

[0111] (2) Location data: GPS positioning technology is used to obtain the location information of escort vehicles in real time. Combined with the Geographic Information System (GIS), the driving route and real-time location of escort vehicles can be accurately tracked so that the monitoring center can keep track of the escort progress at any time and promptly detect abnormal situations, such as deviation from the scheduled route.

[0112] (3) Luggage data: Sensors and NFC identification chips are installed in the escorted cash boxes. The robot collects the unique ID of the luggage, as well as the weight, temperature, humidity and other status data of the luggage, and whether it has been illegally opened, to ensure the safety and integrity of the luggage during the escort process.

[0113] (4) Visual data: When arriving at the escort handover site, such as the entrance of a safe or a bank branch, the robot’s high-definition camera collects the facial information of the warehouse guards and the site environment information in real time to register the escort receipt behavior.

[0114] 2. Financial escort vertical model

[0115] (1) Data preprocessing and feature extraction: Preprocess the collected multimodal data to identify and extract valuable features. For example, accurately identify the shape, color, texture and other features of the escorted luggage from the visual image, and extract the unique identity feature vector from the biometric data, providing a basis for subsequent identification and analysis.

[0116] (2) Personnel and luggage identification: Train the classification and identification capabilities of the financial escort vertical model to accurately match the collected biometric features of personnel and visual features of luggage. Through a large amount of training data, it can accurately determine whether the personnel are legal business outlets, vault managers, cashiers and escorts, and whether the escorted luggage is the registered target luggage, effectively preventing the occurrence of false claims and wrong claims.

[0117] (3) Abnormal behavior detection: Based on the learning of a large amount of normal escort behavior data, the behavior of escort personnel is analyzed and monitored in real time to identify abnormal behavior patterns, such as personnel wandering in non-specified areas, bags being moved abnormally, etc. Once an abnormality is found, an alarm is immediately issued to remind the monitoring personnel to deal with it in a timely manner.

[0118] (4) Data prediction: By analyzing historical data, real-time road conditions, weather conditions and other multi-dimensional information of escort routes, we can discover potential patterns and trends, predict possible risk points and congested sections, plan and schedule escort tasks in advance, arrange safer and faster routes for escort tasks, reduce risks and improve efficiency, and improve the efficiency and safety of escort work.

[0119] 3. Business support applications

[0120] (1) Personnel registration module: registers and manages the basic information and biometric information of escort personnel at various positions. When a person joins the company, their biometric features such as face, fingerprint, and iris are collected and bound to their personal identity information and stored in the system database. During the escort mission, the identity of personnel at various positions is verified through visual and biometric technology to ensure that only legitimate personnel are involved in the escort.

[0121] (2) Luggage registration module: Detailed information about the escorted luggage is registered, including the luggage number, type, weight, bank branch, etc., and associated with the corresponding escort task. During the escort process, the luggage is monitored in real time through visual recognition and sensor data to ensure that the status of the luggage is consistent with the registration information.

[0122] (3) Task Scheduling Module: Develop a reasonable escort task plan based on the bank's business needs and the availability of escort resources. Combined with the data analysis and prediction results provided by the financial escort vertical model, optimize and adjust the escort route and time schedule to improve escort efficiency and reduce risks. At the same time, track the execution of escort tasks in real time and handle abnormal situations in tasks in a timely manner.

[0123] (4) Monitoring and alarm module: Receive and display information from multimodal data acquisition equipment in real time, and monitor the escort process in all directions. When the financial escort vertical model detects an abnormal situation, the system automatically sends an alarm signal and pushes relevant information to monitoring personnel and management personnel so that timely measures can be taken to ensure escort safety.

[0124] Data management and query module: responsible for storing, managing and querying various data generated in the system, including personnel information, luggage information, escort mission records, surveillance videos and other data, supporting query and statistical analysis according to different conditions, and providing data support and decision-making basis for escort management.

[0125] In addition, the specific application fields of the present invention are:

[0126] 1. Financial escort company: By integrating multi-modal financial escort vertical large-scale model robots, we provide intelligent solutions for the management and daily operations of professional escort companies.

[0127] 2. Banks and other financial institutions: Integrate relevant data on financial escort, provide data analysis reports and consulting services to banks and other financial institutions, and help financial institutions conduct security risk control.

[0128] In addition, relevant evidence of the technical effects obtained by the embodiments of the present invention:

[0129] The present invention will expand the existing data dimensions of the escort industry, stimulate data utilization efficiency and enhance data relationship mining capabilities through data fusion.

[0130] Through the analysis of related fields, we can find some effect benefits in adjacent fields:

[0131] In the field of transportation, by integrating the Internet and traffic police data, it supports the perception and prediction of traffic flow in different directions at intersections. By mining the deeper relationships between data, such as the relationship between vehicle flow and signal timing, it dynamically adjusts the duration of traffic lights at each intersection, realizes dynamic tuning of traffic signals, and improves road traffic efficiency.

[0132] In the field of logistics, through in-depth mining and analysis of logistics data, route planning and packaging optimization can be carried out more scientifically, logistics operation efficiency can be improved, operating costs can be reduced, data-driven logistics decision optimization can be achieved, and the value of data can be fully utilized.

[0133] It can be foreseen that the application of the present invention will achieve multiple effects in the field of financial escort, such as Figure 5 .

[0134] Improve the dimension of data perception, open up data links with multimodal robots as the center, and reduce data integration costs.

[0135] Stimulate data utilization efficiency, conduct real-time monitoring and scheduling, optimize business processes, and predict and prevent risks through large-model robots. If a vehicle deviates from the scheduled route or a person is in an abnormal state, issue an early warning in a timely manner and make scheduling adjustments.

[0136] Improve data relationship mining capabilities, and through the integration of multimodal financial escort vertical large-scale model robots, mine the correlation between different data and security risks, the correlation between customer data and service quality data, and the correlation between market trend data and its own business development.

[0137] It should be noted that the embodiments of the present invention can be implemented by hardware, software, or a combination of software and hardware. The hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. It can be understood by a person of ordinary skill in the art that the above-mentioned devices and methods can be implemented using computer executable instructions and / or contained in a processor control code, such as a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. Such code is provided on the carrier medium. The device and its modules of the present invention can be implemented by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., can also be implemented by software executed by various types of processors, and can also be implemented by a combination of the above-mentioned hardware circuits and software, such as firmware.

[0138] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with the technical field within the technical scope disclosed by the present invention and within the spirit and principle of the present invention should be covered by the protection scope of the present invention.

Claims

1. A multi-modal financial escort vertical large-scale model robot, characterized by: The multi-modal integrated financial escort vertical large-scale model robot includes: Multimodal perception layer module, data warehouse module, financial escort vertical large model module, business support application module; The multimodal perception layer module is connected with the data warehouse module and the supporting application module to obtain biometric data, location data, luggage data, and visual data; The data warehouse module is connected with the multimodal perception layer module, the financial escort vertical model module, and the business support application module to store data; The large vertical model module of financial escort is connected with the data warehouse module and the business support application module for data preprocessing and feature extraction, personnel and luggage identification, abnormal behavior detection, and data prediction; The business support application module is connected to the multimodal perception layer module, the data warehouse module, and the financial escort vertical model module, and is used for personnel registration, luggage registration, task scheduling, monitoring and alarm, and data management and query.

2. The multi-modal financial escort vertical large-scale model robot as claimed in claim 1 is characterized in that: The multimodal perception layer module: (1) Biometric data: Use fingerprint recognition and iris recognition biometric technology to collect fingerprint and iris information of escort personnel; (2) Location data: GPS positioning technology is used to obtain the location information of the escort vehicle in real time. Combined with the geographic information system, the driving route and real-time location of the escort vehicle can be accurately tracked so that the monitoring center can keep track of the escort progress at any time and detect abnormal situations in a timely manner. (3) Luggage data: Sensors and NFC identification chips are installed in the escorted cash boxes. The robot collects the unique ID of the luggage, as well as the weight, temperature, humidity status data of the luggage, and whether it has been illegally opened, to ensure the safety and integrity of the luggage during the escort process; (4) Visual data: When arriving at the escort handover site, at key locations such as the safe door and bank branch entrance, the robot’s high-definition camera collects the facial information of the vault guards and the site environment information in real time to register the escort receipt behavior.

3. The multi-modal financial escort vertical large-scale model robot as claimed in claim 1 is characterized in that: The financial escort vertical model module: (1) Data preprocessing and feature extraction: Preprocess the collected multimodal data to identify and extract valuable features; (2) Personnel and luggage identification: Train the classification and recognition capabilities of the financial escort vertical large model to accurately match the collected biometric features of personnel and visual features of luggage. Through a large amount of training data, accurately determine whether the personnel are legal business outlets’ warehouse managers, vault managers, cash delivery personnel, and escorts, and whether the escorted luggage is the registered target luggage, effectively preventing the occurrence of false or wrong collection. (3) Abnormal behavior detection: Based on the learning of a large amount of normal escort behavior data, the escort personnel’s behavior and actions are analyzed and monitored in real time to identify abnormal behavior patterns; (4) Data prediction: By analyzing the historical data, real-time road conditions, and weather conditions of the escort routes, we can discover the underlying patterns and trends, predict possible risk points and congested sections, and plan and schedule escort tasks in advance.

4. The multi-modal financial escort vertical large-scale model robot as claimed in claim 1 is characterized in that: The business support application module: (1) Personnel registration module: registers and manages the basic information and biometric information of escort personnel at various positions; when personnel join the service, their facial, fingerprint, and iris biometrics are collected and bound to their personal identity information and stored in the system database; during escort missions, identity verification of personnel at various positions is performed through visual and biometric recognition technology to ensure that only legitimate personnel participate in escort missions; (2) Luggage registration module: Detailed information about the escorted luggage is registered, including the luggage number, type, weight, and bank branch information, and associated with the corresponding escort task; during the escort process, the luggage is monitored in real time through visual recognition and sensor data to ensure that the status of the luggage is consistent with the registration information; (3) Task Scheduling Module: Develop a reasonable escort task plan based on the bank’s business needs and the availability of escort resources; optimize and adjust escort routes and time arrangements based on the data analysis and prediction results provided by the financial escort vertical model; (4) Monitoring and alarm module; (5)Data management and query module.

5. The multi-modal financial escort vertical large-scale model robot as claimed in claim 4 is characterized in that: The monitoring and alarm module receives and displays information from multimodal data acquisition equipment in real time, and monitors the escort process in all directions. When the financial escort vertical model detects an abnormal situation, the system automatically sends an alarm signal and pushes relevant information to monitoring personnel and managers.

6. The multi-modal financial escort vertical large-scale model robot as claimed in claim 4 is characterized in that: The data management and query module is responsible for storing, managing and querying various types of data generated in the system, including personnel information, luggage information, escort mission records, and surveillance video data. It supports query and statistical analysis according to different conditions, and provides data support and decision-making basis for escort management.

7. A control method for a fusion multi-modal financial escort vertical large-scale model robot implementing the fusion multi-modal financial escort vertical large-scale model robot as described in any one of claims 1 to 6, characterized in that: The multi-modal integrated financial escort vertical large-scale model robot control method includes: Step 1, obtaining biometric data, location data, luggage data, and visual data through a multimodal perception layer module; Step 2, storing data through the data warehouse module; Step 3: Preprocess data and extract features through the large model module of financial escort vertical category, identify personnel and luggage, detect abnormal behavior, and predict data; Step 4: Use the business support application modules for personnel registration, luggage registration, task scheduling, monitoring and alarm, and data management and query.

8. A computer device, characterized in that: The computer device includes a memory and a processor, the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the multi-modal financial escort vertical large model robot control method as described in claim 7.

9. A computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor executes the steps of the multi-modal financial escort vertical large-scale model robot control method as described in claim 7.

10. An information data processing terminal, characterized in that: The information data processing terminal is used to realize the large-scale vertical model robot of financial escort integrating multiple modes as described in any one of claims 1-6.

Images